Incomplete data ensemble classification using imputation-revision framework with local spatial neighborhood information

Estimation of missing values in astronomical survey data: An improved local approach using cluster directed neighbor selection

“…Besides the analysis of sky survey data, these new methods are to be assessed with datasets belong to other problem domains like bioinformatics. At last, possible applications of imputation techniques to fuzzy reasoning [43], analysis of mixed-type data [44] and multiple-imputation modeling [18] can also be further studied.…”

“…Apart from these, a rich collection of learning based or data driven models is proposed in the literature [17]. One group of these focuses on the concept of multiple imputation that exploits a set of techniques to produce different estimates for a missing value, while the other called single imputation generates one guess for any missing entries [18]. In general, there are three major families of single imputation techniques: knowledge-assisted, global and local categories, respectively.…”

Evidence Combination for Pattern Classification With Missing Data

“…In the traditional classification methods, like SVM, decision tree, the training pattern and test pattern are generally complete, and the missing values are not considered. Many classification methods have emerged to address the issue of incomplete patterns [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27]. These methods can be broadly divided into four categories: removing incomplete patterns, model-based classification, machine learning-based imputation methods, and direct classification of incomplete pattern.…”